Research And Application Of Poverty Return Prediction Model Based On Data Mining

Accurate poverty alleviation is an important strategic measure to build a well-off society in an all-round way in China.While remarkable achievements have been made in poverty alleviation in China,there is a fluctuating trend in poverty alleviation achievements in many poverty-stricken areas.Because of the phenomenon of poverty-returning,poverty-free people return to poverty,which is a difficult problem to be solved urgently in the final stage of our country's poverty alleviation and is also a severe challenge for our country after building a well-off society in an all-round way.Therefore,It is helpful for Poverty Alleviation staffs to predict the future situation of poverty-free people poverty-returning by mining the characteristics of poverty-returning and constructing the prediction model of poverty-returning,and to formulate targeted and more effective assistance measures in advance,which has important application value for the development of anti-poverty strategy in China.In order to consolidate the achievements of poverty alleviation and reduce the rate of poverty-returning,this paper constructs a prediction model of poverty-returning based on the current poverty situation in China,using the advantages of neural network with strong non-linear input-output mapping ability,and provides scientific and effective decision support for poverty alleviation staffs.The specific contents of this paper are as follows:(1)Analysis of poverty-returning Event Association Rules Based on association rule algorithm.According to SPCA model,the influencing factors of poverty-returning events are selected,and the database of poverty-returning events is established by generalizing the influencing factors of data attributes.The influencing factors of poverty-returning are analyzed by FP-Growth algorithm,and the maximum poverty-returning feature set and its association rules are mined.(2)Prediction model of poverty-returning based on BP neural network is constructed.Based on the event database of poverty alleviation,this paper quantifies the feature item set.trains the neural network with Bayesian regularization algorithm,constructs the prediction model of poverty alleviation based on BP neural network,and validates the prediction model of poverty alleviation.It can locate the poverty-free people with great possibility of poverty alleviation,and provide early warning for the poverty alleviation workers so as to facilitate the timely adjustment and assistance of the poverty alleviation staffs.In order to reduce the rate of poverty-returning,the support plan should prevent incidents of poverty-returning in advance.(3)The design and implementation of poverty prediction and analysis system.Based on the above models and methods,combined with the needs of users in the field of poverty alleviation,this paper designs and implements a decision-making and analysis system for poverty alleviation prediction.Its functions include multi-condition query,detailed information of farmers,dynamic statistics,association irule mining and poverty alleviation prediction analysis.In order to make the poverty alleviation staffs better use the system to carry out poverty alleviation work,the predicted results of association rules and neural networks are further applied to Poverty Alleviation Policies and specific implementation.